Input method word frequency adjustment method and device

Abstract

The present application discloses an input method word frequency adjustment method and device, which are used to solve the technical problem of poor word frequency adjustment effect of input method phrases. An input method word frequency adjustment method includes the following steps: obtaining corpus data; segmenting the corpus data through a word segmentation model to generate a number of word segmentation units; annotating the word segmentation units through a phonetic recognition model to generate word segmentation unit syllables; saving word segmentation units with the same syllables to the same syllable vocabulary; counting the occurrence probability of the first word segmentation unit in the same syllable vocabulary; comparing the occurrence probability of the first word segmentation unit with a preset threshold to obtain a comparison result; adjusting the word frequency of the first word segmentation unit according to the comparison result; arranging the word segmentation unit order of the syllable vocabulary where the first word segmentation unit is located in a preset order according to the adjusted word frequency of the first word segmentation unit, and updating the syllable vocabulary. By dynamically adjusting the word frequency of phrases in the same syllable vocabulary, the accuracy of input is improved.

Description

Technical Field

[0001] The present application relates to the field of input technology, and in particular to an input method word frequency adjustment method and device. Background Art

[0002] With the continuous development and advancement of social economy, culture, and technology, the vocabulary used in daily life and in professional fields is also constantly evolving. This has led to the rapid emergence and iteration of many new words. To adapt to the emergence of new words, input methods need to promptly adjust the frequency of phrases in the input method. Effective adjustment of input method frequency can greatly improve the accuracy and timeliness of input methods.

[0003] In the process of implementing the prior art, the inventors found that:

[0004] In existing input methods, word frequencies are often modified by using a phrase coefficient. This approach lacks an effective adjustment mechanism for new phrases. This is because the method is very sensitive to initial word frequencies; either too low or too high initial frequencies are detrimental to the ranking of new words in the vocabulary. Furthermore, over time, the frequency of the original phrases no longer aligns with current user input habits, thus affecting input accuracy.

[0005] Therefore, it is necessary to provide an input method word frequency adjustment method and device to solve the technical problem of poor input method phrase frequency adjustment effect. Summary of the Invention

[0006] The embodiments of the present application provide an input method word frequency adjustment method and device to solve the technical problem of poor input method phrase frequency adjustment effect.

[0007] Specifically, a method for adjusting word frequency of an input method includes the following steps:

[0008] Get corpus data;

[0009] Segmenting the corpus data using a word segmentation model to generate a number of word segmentation units;

[0010] Pronunciation is performed on the word segmentation unit through a phonetic recognition model to generate word segmentation unit syllables;

[0011] Save the word units with the same syllables into the same syllable vocabulary;

[0012] Count the occurrence probability of the first word segmentation unit in the same syllable vocabulary;

[0013] Comparing the occurrence probability of the first word segmentation unit with a preset threshold to obtain a comparison result;

[0014] According to the comparison result, the word frequency of the first word segmentation unit is adjusted;

[0015] According to the adjusted word frequency of the first word segmentation unit, the word segmentation unit sequence of the syllable vocabulary where the first word segmentation unit is located is arranged in a preset order, and the syllable vocabulary is updated.

[0016] Furthermore, based on the comparison result, the word frequency of the first word segmentation unit is adjusted, specifically including:

[0017] When the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, increasing the word frequency of the first word segmentation unit;

[0018] When the occurrence probability of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, the word frequency of the first word segmentation unit is weakened.

[0019] Furthermore, when the occurrence probability of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, the word frequency of the first word segmentation unit is enhanced, specifically including:

[0020] According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows:

[0021]

[0022] Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first participle unit In the same syllable Probability of occurrence in the vocabulary;

[0023] According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows:

[0024]

[0025] Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0026] According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows:

[0027]

[0028] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

[0029] Furthermore, when the occurrence probability of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, the word frequency of the first word segmentation unit is weakened, specifically including:

[0030] According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows:

[0031]

[0032] Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit In the same syllable Probability of occurrence in the vocabulary;

[0033] According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows:

[0034]

[0035] Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0036] According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows:

[0037]

[0038] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

[0039] Furthermore, the word segmentation model is a neural network algorithm model based on language grammatical rules, which is used to identify the grammatical components of the corpus data or the parts of speech of the corpus data phrases, and to segment the corpus data according to the grammatical components or the parts of speech of the corpus data phrases;

[0040] The phonetic recognition model is a neural network algorithm model based on language grammar rules, which is used to recognize syllables in corpus data and perform phonetic annotation.

[0041] The embodiment of the present application also provides an input method word frequency adjustment device.

[0042] Specifically, an input method word frequency adjustment device includes:

[0043] Acquisition module, used to obtain corpus data;

[0044] A word segmentation module, used to segment the corpus data using a word segmentation model to generate a number of word segmentation units;

[0045] A phonetic notation module is used to perform phonetic notation on the word segmentation unit through a phonetic recognition model to generate a word segmentation unit syllable;

[0046] A storage module, used for storing word segmentation units with the same syllables into the same syllable vocabulary;

[0047] A statistical module, used to count the occurrence probability of the first word segmentation unit in the same syllable vocabulary;

[0048] A comparison module, configured to compare the occurrence probability of the first word segmentation unit with a preset threshold to obtain a comparison result;

[0049] The adjustment module is used to adjust the word frequency of the first word segmentation unit according to the comparison result; and is used to arrange the word segmentation unit order of the syllable vocabulary where the first word segmentation unit is located in a preset order according to the word frequency of the adjusted first word segmentation unit, and update the syllable vocabulary.

[0050] Furthermore, the adjustment module is used to adjust the word frequency of the first word segmentation unit according to the comparison result, specifically for:

[0051] When the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, increasing the word frequency of the first word segmentation unit;

[0052] When the occurrence probability of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, the word frequency of the first word segmentation unit is weakened.

[0053] Furthermore, the adjustment module is configured to enhance the word frequency of the first word segmentation unit when the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, specifically for:

[0054] According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows:

[0055]

[0056] Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit In the same syllable Probability of occurrence in the vocabulary;

[0057] According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows:

[0058]

[0059] Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0060] According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows:

[0061]

[0062] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

[0063] Furthermore, the adjustment module is configured to weaken the word frequency of the first word segmentation unit when the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, specifically for:

[0064] According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows:

[0065]

[0066] Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit Probability of occurrence in the same syllable vocabulary;

[0067] According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows:

[0068]

[0069] Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0070] According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows:

[0071]

[0072] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

[0073] Furthermore, the word segmentation model is a neural network algorithm model based on language grammatical rules, which is used to identify the grammatical components of the corpus data or the parts of speech of the corpus data phrases, and to segment the corpus data according to the grammatical components or the parts of speech of the corpus data phrases;

[0074] The phonetic recognition model is a neural network algorithm model based on language grammar rules, which is used to recognize syllables in corpus data and perform phonetic annotation.

[0075] The technical solutions provided in the embodiments of the present application have at least the following beneficial effects:

[0076] By dynamically adjusting the frequency of phrases in the input method, the vocabulary can adapt to the addition of new words or changes in the frequency of existing phrases, thereby improving the accuracy of input. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0078] Figure 1 A flowchart of an input method word frequency adjustment method provided in an embodiment of the present application.

[0079] Figure 2 A schematic diagram of the structure of an input method word frequency adjustment device provided in an embodiment of the present application.

[0080] 100 input method word frequency adjustment device

[0081] 11 Get Module

[0082] 12 word segmentation modules

[0083] 13 Zhuyin module

[0084] 14 storage modules

[0085] 15 Statistics Module

[0086] 16 comparison modules

[0087] 17 Adjustment Module DETAILED DESCRIPTION

[0088] To make the purpose, technical solutions, and advantages of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the specific embodiments of this application and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0089] The vocabulary people use daily is constantly evolving, leading to the rapid emergence and iteration of many new words. Consequently, the words people frequently input are also constantly changing. To adapt to the emergence of new words, input methods need to promptly adjust the frequency of phrases in the input method. Effectively adjusting the frequency of phrases in the input method can greatly improve the accuracy and timeliness of the input method.

[0090] In existing input methods, word frequencies are often modified by using a phrase coefficient. This approach lacks an effective adjustment mechanism for new phrases. This is because the method is very sensitive to initial word frequencies; either too low or too high initial frequencies are detrimental to the ranking of new words in the vocabulary. Furthermore, over time, the frequency of the original phrases no longer aligns with current user input habits, thus affecting input accuracy.

[0091] Therefore, the inventor provides an input method word frequency adjustment method and device to solve the technical problem of poor input method phrase frequency adjustment effect.

[0092] Please refer to Figure 1 This application discloses an input method word frequency adjustment method, comprising the following steps:

[0093] S110: Acquire corpus data.

[0094] It is understood that the corpus data can be understood as at least one text combination of characters, words, sentences, and documents. For example, video titles on video websites, comments on video websites, documents on information websites, etc. can all be the corpus data described in this application document.

[0095] S120: Segmenting the corpus data using a word segmentation model to generate a number of word segmentation units.

[0096] It is understood that the word segmentation model can be a neural network algorithm model based on language grammatical rules, which is used to identify the grammatical components of corpus data or the parts of speech of corpus data phrases, and segment the corpus data according to the grammatical components or the parts of speech of corpus data phrases to generate a number of word segmentation units. The grammatical components of the corpus data can be expressed as grammatical components such as the subject, predicate, object, verb, attributive, adverbial, complement, predicative, and appositive in a sentence. The parts of speech of the corpus data phrases can be expressed as phrase parts of speech such as nouns, pronouns, verbs, prepositions, and numerals.

[0097] Therefore, step S120 can be expressed as follows: inputting corpus data into the word segmentation model; then the word segmentation model segments the corpus data according to the parts of speech of phrases such as nouns, pronouns, verbs, prepositions, and numerals in the sentence; finally, the word segmentation model splits the corpus data into several word segmentation units.

[0098] Alternatively, step S120 can also be performed as follows: inputting corpus data into a word segmentation model; then the word segmentation model segments the corpus data according to grammatical components such as the subject, predicate, object, verb, attributive, adverbial, complement, predicative, appositive, etc. in the sentence; finally, the word segmentation model splits the corpus data into several word segmentation units.

[0099] S130: annotating the word segmentation unit with the phonetic recognition model to generate word segmentation unit syllables.

[0100] It can be understood that the syllable recognition model can be a neural network algorithm model based on language grammar rules for recognizing the syllables of the word segmentation units. Specifically, the syllables can be understood as the phonetic notations of the text, such as International Phonetic Alphabet, Chinese pinyin, etc.

[0101] S140: Save the word segmentation units with the same syllables into the same syllable word library.

[0102] It should be noted that the word library can be understood as a database storing word segmentation units with the same syllable. Because there are homophonic phenomena in the text, there will be multiple corresponding texts under the same syllable. Therefore, the syllables of the word segmentation units in the same word library are the same. It should also be noted that the word library stores at least one word segmentation unit.

[0103] S150: Statistically analyze the occurrence probability of the first word segmentation unit in the same syllable word library.

[0104] It can be understood that at least one word segmentation unit is stored in the same word library. For the sake of concise description, only an arbitrary word segmentation unit in the same word library is described here, and it should not be construed as a limitation on the scope of the invention patent.

[0105] For the sake of easy representation, an arbitrary word segmentation unit in the same word library is denoted as the first word segmentation unit here. Then the occurrence probability of the first word segmentation unit can be understood as a parameter describing the relationship between the occurrence frequency of the first word segmentation unit and the occurrence frequencies of all word groups in the same word library.

[0106] For example, in the word library with the syllable xin’guan, there are word segmentation units such as "新馆" (new museum), "信管" (fuse), "新官" (new official), "心管" (cardiovascular), "新灌" (new irrigation), etc. Statistically analyze the occurrence frequencies of each word segmentation unit in the word library with the syllable xin’guan. Through a preset calculation matrix for the occurrence probability of word segmentation units, the occurrence probability of the word segmentation unit "新灌" (new irrigation) can be calculated as P(新灌|xin’guan)=0.001836; the occurrence probability of the word segmentation unit "信管" (fuse) is P(信管|xin’guan)=0.000024.

[0107] S160: Compare the occurrence probability of the first word segmentation unit with a preset threshold to obtain a comparison result.

[0108] It should be noted that when comparing the occurrence probability of the first word segmentation unit with a preset threshold, there will be two situations for the obtained comparison result, that is, the occurrence probability of the first word segmentation unit in the same syllable word library is greater than or equal to the preset threshold; or the occurrence probability of the first word segmentation unit in the same syllable word library is less than the preset threshold. And the comparison result will determine the word frequency adjustment method of the first word segmentation unit.

[0109] S170: Adjust the word frequency of the first word segmentation unit according to the comparison result.

[0110] It should be noted that when the occurrence probability of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, the word frequency of the first word segmentation unit is enhanced.

[0111] When the occurrence probability of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, the word frequency of the first word segmentation unit is weakened.

[0112] Specifically, when the probability of the first word segmentation unit appearing in the same syllable vocabulary is greater than or equal to a preset threshold, the system calculates the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment based on the probability of the first word segmentation unit appearing in the same syllable vocabulary; it is expressed as follows:

[0113]

[0114] Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first participle unit In the same syllable Probability of occurrence in the vocabulary;

[0115] According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows:

[0116]

[0117] Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0118] According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows:

[0119]

[0120] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

[0121] It should be emphasized that the probability of occurrence of the first segmented word unit in the same syllable vocabulary is a statistical result of the system within a preset time period, which facilitates dynamic adjustment of the input method phrase frequency. This dynamic adjustment of the input method phrase frequency is manifested by the system adjusting the phrase frequency in the same syllable vocabulary at preset intervals, specifically the Kth round of adjustment and the K+1th round of adjustment mentioned above.

[0122] When the probability of the first word segmentation unit appearing in the same syllable vocabulary is less than a preset threshold, the system will calculate the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment based on the probability of the first word segmentation unit appearing in the same syllable vocabulary; it is expressed as follows:

[0123]

[0124] Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit In the same syllable Probability of occurrence in the vocabulary;

[0125] According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows:

[0126]

[0127] Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0128] According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows:

[0129]

[0130] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of 。

[0131] It should be emphasized that the occurrence probability of the first segmentation unit in the above-mentioned same-syllable vocabulary is the statistical result of the system within a preset time, so as to facilitate the dynamic adjustment of the word frequency of the input method. And the dynamic adjustment of the word frequency of the input method is manifested as that the system adjusts the word frequency of the phrases in the same-syllable vocabulary every preset time, specifically manifested as the above-mentioned K-th round of adjustment and the (K + 1)-th round of adjustment.

[0132] S180: According to the adjusted word frequency of the first segmentation unit, arrange the segmentation unit order of the syllable vocabulary where the first segmentation unit is located in a preset order, and update the syllable vocabulary.

[0133] It should be noted that the order of arranging the segmentation units of the syllable vocabulary where the first segmentation unit is located in a preset order can be that the system rearranges the segmentation units in the vocabulary in descending order of the word frequency of the phrases according to the adjusted word frequency of the first segmentation unit.

[0134] In a specific implementation manner provided by the present application, after the system obtains a large amount of corpus data, through a segmentation model, the corpus data is segmented to generate a number of segmentation units. And through a phonetic recognition model, the segmentation units are phonetically annotated to generate the syllables of the segmentation units. Then, the segmentation units with the same syllables are saved to the same syllable vocabulary.

[0135] Among them, in the vocabulary with the syllable xin’guan, there are segmentation units such as "新馆" (new museum), "信管" (letter tube), "新官" (new official), "心管" (heart tube), "新灌" (new irrigation), etc. Count the occurrence frequencies of each segmentation unit in the vocabulary with the syllable xin’guan. Through a preset calculation matrix of the occurrence probability of the segmentation unit, the occurrence probability of the segmentation unit "新灌" (new irrigation) can be calculated as P(新灌|xin’guan)=0.001836; the occurrence probability of the segmentation unit "信管" (letter tube) is P(信管|xin’guan)=0.000024.

[0136] Compare P(信管|xin’guan)=0.000024 and P(新灌|xin’guan)=0.001836 with a preset threshold hit respectively. The comparison result is that P(新灌|xin’guan) is greater than the set threshold, while P(信管|xin’guan) is less than the threshold. Then it is necessary to calculate the enhanced word frequency for "新灌" (new irrigation) and calculate the weakened word frequency for "信管" (letter tube).

[0137] Substitute P(信管|xin’guan)=0.000024 into the calculation equation in step S170, and the adjusted word frequency of "信管" (letter tube) can be obtained as 169. Substitute P(新灌|xin’guan)=0.001836 into the calculation equation in step S170, and the adjusted word frequency of "新灌" (new irrigation) can be obtained as 171.

[0138] Finally, the system arranges the adjusted word segmentation units in descending order of word frequency. In the updated same-syllable word library, the word segmentation unit "Xinguàn" ranks first, and the word segmentation unit "Xìnguǎn" ranks second.

[0139] During this work, the input method word frequency adjustment method dynamically adjusts the word frequency of phrases in the same-syllable word library, so that the word library can adapt to the addition of new words or the change of the word frequency of original phrases, improving the input accuracy.

[0140] Please refer to Figure 2 , to support the input method word frequency adjustment method, this application also provides an input method word frequency adjustment device 100, including:

[0141] An acquisition module 11, configured to acquire corpus data;

[0142] A word segmentation module 12, configured to segment the corpus data through a word segmentation model to generate a number of word segmentation units;

[0143] A phonetic notation module 13, configured to phonetically annotate the word segmentation units through a phonetic recognition model to generate word segmentation unit syllables;

[0144] A storage module 14, configured to save word segmentation units with the same syllable to the same-syllable word library;

[0145] A statistics module 15, configured to count the occurrence probability of the first word segmentation unit in the same-syllable word library;

[0146] A comparison module 16, configured to compare the occurrence probability of the first word segmentation unit with a preset threshold to obtain a comparison result;

[0147] An adjustment module 17, configured to adjust the word frequency of the first word segmentation unit according to the comparison result; configured to arrange the word segmentation unit order of the syllable word library where the first word segmentation unit is located in a preset order according to the adjusted word frequency of the first word segmentation unit, and update the syllable word library.

[0148] It can be understood that the corpus data acquired by the acquisition module 11 can be understood as including at least one text combination of characters, words, sentences, and documents. For example, data such as video titles on video websites, comments on video websites, and documents on data websites can all be the corpus data described in this application document.

[0149] The word segmentation model called by the word segmentation module 12 can be a neural network algorithm model based on language grammatical rules, which is used to identify the grammatical components of the corpus data or the word groups of the corpus data, and segment the corpus data according to the grammatical components or the word groups of the corpus data to generate a number of word segmentation units. The grammatical components of the corpus data can be expressed as grammatical components such as the subject, predicate, object, verb, attributive, adverbial, complement, predicative, and appositive in a sentence. The word groups of the corpus data can be expressed as word groups such as nouns, pronouns, verbs, prepositions, and numerals.

[0150] Therefore, the word segmentation model can segment the corpus data according to the word groups such as nouns, pronouns, verbs, prepositions, and numerals in the sentence; finally, the word segmentation model splits the corpus data into several word segmentation units.

[0151] Alternatively, the word segmentation model can also segment the corpus data according to grammatical components such as the subject, predicate, object, verb, attributive, adverbial, complement, predicative, appositive, etc. in the sentence; finally, the word segmentation model splits the corpus data into several word segmentation units.

[0152] The phonetic recognition model called by the phonetic module 13 can be a neural network algorithm model based on language grammar rules, which is used to recognize the syllables of the word segmentation unit. Specifically, the syllables can be understood as the phonetic symbols of the text, such as the International Phonetic Alphabet, Pinyin and other phonetic symbols.

[0153] The storage module 14 is used to store segmentation units with the same syllable in the same syllable vocabulary. The vocabulary can be understood as a database that stores segmentation units with the same syllable. Because texts have homophones, there will be multiple corresponding texts under the same syllable. Therefore, the segmentation units in the same vocabulary have the same syllable. It should also be noted that the vocabulary stores at least one segmentation unit.

[0154] It is understandable that, since at least one word segmentation unit is stored in the same word library, for the sake of simplicity, the statistical module 15 will describe the word frequency adjustment based on any word segmentation unit in the same word library, and this should not be understood as limiting the scope of the invention patent.

[0155] For ease of representation, the statistical module 15 records any word segmentation unit in the same vocabulary as the first word segmentation unit. The occurrence probability of the first word segmentation unit can be understood as a parameter describing the relationship between the occurrence frequency of the first word segmentation unit and the occurrence frequency of all phrases in the same vocabulary.

[0156] For example, in the word library with the syllable "xin’guan", there are word segmentation units such as "新馆", "信管", "新官", "心管", "新灌", etc. By counting the occurrence frequencies of each word segmentation unit in the word library with the syllable "xin’guan", through a pre-set calculation matrix of the occurrence probabilities of word segmentation units, the occurrence probability of the word segmentation unit "新灌" can be calculated as P(新灌|xin’guan)=0.001836; the occurrence probability of the word segmentation unit "信管" is P(信管|xin’guan)=0.000024.

[0157] The comparison module 16 is used to compare the occurrence probability of the first word segmentation unit with a pre-set threshold to obtain a comparison result. When comparing the occurrence probability of the first word segmentation unit with the pre-set threshold, two situations may occur in the obtained comparison result, that is, the occurrence probability of the first word segmentation unit in the same syllable word library is greater than or equal to the pre-set threshold; or the occurrence probability of the first word segmentation unit in the same syllable word library is less than the pre-set threshold. And the comparison result will determine the word frequency adjustment method of the first word segmentation unit.

[0158] The adjustment module 17 is used to adjust the word frequency of the first word segmentation unit according to the comparison result. Specifically, when the occurrence probability of the first word segmentation unit in the same syllable word library is greater than or equal to the pre-set threshold, enhance the word frequency of the first word segmentation unit. When the occurrence probability of the first word segmentation unit in the same syllable word library is less than the pre-set threshold, weaken the word frequency of the first word segmentation unit.

[0159] Specifically, when the occurrence probability of the first word segmentation unit in the same syllable word library is greater than or equal to the pre-set threshold, the adjustment module 17 calculates the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment according to the occurrence probability of the first word segmentation unit in the same syllable word library; it is expressed as follows:

[0160]

[0161] In the formula, is the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, is the first word segmentation unit in the same syllable word library occurrence probability;

[0162] The adjustment module 17 calculates the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment according to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment; it is expressed as follows:

[0163]

[0164] In the formula, is the first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0165] The adjustment module 17 calculates the word frequency of the first word segmentation unit in the K+1th round of adjustment based on the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment; which is expressed as follows:

[0166]

[0167] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

[0168] It should be emphasized that the probability of occurrence of the first segmented word unit in the same syllable vocabulary described above is the statistical result of the adjustment module 17 within a preset time period, so as to facilitate dynamic adjustment of the input method phrase frequency. The dynamic adjustment of the input method phrase frequency is manifested by the adjustment module 17 adjusting the phrase frequency in the same syllable vocabulary at preset intervals, specifically the Kth round of adjustment and the K+1th round of adjustment described above.

[0169] When the probability of the first word segmentation unit appearing in the same syllable vocabulary is less than a preset threshold, the adjustment module 17 calculates the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment based on the probability of the first word segmentation unit appearing in the same syllable vocabulary; it is expressed as follows:

[0170]

[0171] Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit In the same syllable Probability of occurrence in the vocabulary;

[0172] The adjustment module 17 calculates the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment based on the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment; which is expressed as follows:

[0173]

[0174] Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment;

[0175] The adjustment module 17 calculates the word frequency of the first word segmentation unit in the K+1th round of adjustment based on the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment; which is expressed as follows:

[0176]

[0177] Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

[0178] It should be emphasized that the probability of occurrence of the first segmented word unit in the same syllable vocabulary described above is the statistical result of the adjustment module 17 within a preset time period, so as to facilitate dynamic adjustment of the input method phrase frequency. The dynamic adjustment of the input method phrase frequency is manifested by the adjustment module 17 adjusting the phrase frequency in the same syllable vocabulary at preset intervals, specifically the Kth round of adjustment and the K+1th round of adjustment described above.

[0179] The adjustment module 17 is further configured to arrange the order of the segmentation units in the syllable lexicon where the first segmentation unit is located in a preset order based on the word frequency of the adjusted first segmentation unit, thereby updating the syllable lexicon. Specifically, arranging the order of the segmentation units in the syllable lexicon where the first segmentation unit is located in a preset order may be that the adjustment module 17 rearranges the segmentation units in the lexicon in descending order of phrase frequency based on the word frequency of the adjusted first segmentation unit.

[0180] In a specific embodiment provided herein, the acquisition module 11 of the input method word frequency adjustment device 100 acquires a large amount of corpus data. The word segmentation module 12 segments the corpus data using a word segmentation model to generate a plurality of word segmentation units. The phonetic notation module 13 then annotates the word segmentation units using a phonetic recognition model to generate syllables for the word segmentation units. The storage module 14 stores word segmentation units with the same syllables in the same syllable database.

[0181] Among them, in the word library with the syllable "xin’guan", there are word segmentation units such as "新馆", "信管", "新官", "心管", "新灌", etc. The statistical module 15 counts the occurrence frequencies of each word segmentation unit in the word library with the syllable "xin’guan". Through a preset calculation matrix of the occurrence probabilities of word segmentation units, the occurrence probability of the word segmentation unit "新灌" can be calculated as P(新灌|xin’guan) = 0.001836; the occurrence probability of the word segmentation unit "信管" is P(信管|xin’guan) = 0.000024.

[0182] The comparison module 16 compares P(信管|xin’guan) = 0.000024 and P(新灌|xin’guan) = 0.001836 with a preset threshold hit respectively. The comparison result is that P(新灌|xin’guan) is greater than the set threshold, while P(信管|xin’guan) is less than the threshold. Then it is necessary to calculate the enhanced word frequency for "新灌" and the weakened word frequency for "信管".

[0183] The adjustment module 17 substitutes P(信管|xin’guan) = 0.000024 into the above calculation equation of the enhanced word frequency, and the adjusted word frequency of "信管" can be obtained as 169. Substituting P(新灌|xin’guan) = 0.001836 into the above calculation equation of the weakened word frequency, the adjusted word frequency of "新灌" can be obtained as 171.

[0184] Finally, the adjustment module 17 arranges the adjusted word segmentation units in descending order of word frequency. In the updated word library of the same syllable, the word segmentation unit "新灌" ranks first, and the word segmentation unit "信管" ranks second.

[0185] During this work, the input method word frequency adjustment device 100 dynamically adjusts the word frequencies of phrases in the word library of the same syllable, so that the word library can adapt to the addition of new words or the change of the word frequencies of original phrases, improving the input accuracy.

[0186] It should be noted that the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to this process, method, commodity or device. Without more limitations, an element defined by the statement "including a..." does not exclude the existence of another identical element in the process, method, commodity or device including the said element.

[0187] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for adjusting word frequency of an input method, characterized in that: The following steps are involved: Get corpus data; Segmenting the corpus data using a word segmentation model to generate a number of word segmentation units; Pronunciation is performed on the word segmentation unit through a phonetic recognition model to generate word segmentation unit syllables; Save the word units with the same syllables into the same syllable vocabulary; Count the occurrence probability of the first word segmentation unit in the same syllable vocabulary; Comparing the occurrence probability of the first word segmentation unit with a preset threshold to obtain a comparison result; According to the comparison result, the word frequency of the first word segmentation unit is adjusted; According to the adjusted word frequency of the first word segmentation unit, the word segmentation unit order of the syllable word library where the first word segmentation unit is located is arranged in a preset order, and the syllable word library is updated; According to the comparison result, the word frequency of the first word segmentation unit is adjusted, specifically including: When the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, increasing the word frequency of the first word segmentation unit; When the occurrence probability of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, weakening the word frequency of the first word segmentation unit; When the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, the word frequency of the first word segmentation unit is enhanced, specifically including: According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows: ， Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first participle unit In the same syllable Probability of occurrence in the vocabulary; According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows: ， Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment; According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows: ， Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

2. The input method word frequency adjustment method according to claim 1, wherein: When the occurrence probability of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, weakening the word frequency of the first word segmentation unit specifically includes: According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows: ， Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit In the same syllable Probability of occurrence in the vocabulary; According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows: ， Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment; According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows: ， Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

3. The input method word frequency adjustment method according to claim 1, wherein: The word segmentation model is a neural network algorithm model based on language grammar rules, which is used to identify the grammatical components of corpus data or the parts of speech of corpus data phrases, and segment the corpus data according to the grammatical components or the parts of speech of corpus data phrases; The phonetic recognition model is a neural network algorithm model based on language grammar rules, which is used to recognize syllables in corpus data and perform phonetic annotation.

4. An input method word frequency adjustment device, characterized in that: include: Acquisition module, used to obtain corpus data; A word segmentation module, used to segment the corpus data using a word segmentation model to generate a number of word segmentation units; A phonetic notation module is used to perform phonetic notation on the word segmentation unit through a phonetic recognition model to generate a word segmentation unit syllable; A storage module, used for storing word segmentation units with the same syllables into the same syllable vocabulary; A statistical module, used to count the occurrence probability of the first word segmentation unit in the same syllable vocabulary; A comparison module, configured to compare the occurrence probability of the first word segmentation unit with a preset threshold to obtain a comparison result; An adjustment module, configured to adjust the word frequency of the first word segmentation unit according to the comparison result; for arranging the order of the word segmentation units of the syllable vocabulary where the first word segmentation unit is located in a preset order according to the word frequency of the first word segmentation unit after adjustment, and updating the syllable vocabulary; The adjustment module is used to adjust the word frequency of the first word segmentation unit according to the comparison result, specifically to: When the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, increasing the word frequency of the first word segmentation unit; When the occurrence probability of the first word segmentation unit in the same syllable vocabulary is less than a preset threshold, weakening the word frequency of the first word segmentation unit; The adjustment module is configured to enhance the word frequency of the first word segmentation unit when the probability of occurrence of the first word segmentation unit in the same syllable vocabulary is greater than or equal to a preset threshold, specifically for: According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows: ， Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit In the same syllable Probability of occurrence in the vocabulary; According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows: ， Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment; According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows: ， Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

5. The input method word frequency adjustment device according to claim 4, characterized in that: The adjustment module is configured to weaken the word frequency of the first word segmentation unit when the probability of the first word segmentation unit appearing in the same syllable vocabulary is less than a preset threshold, specifically for: According to the occurrence probability of the first word segmentation unit in the same syllable vocabulary, the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; it is expressed as follows: ， Where, The first word segmentation unit The adjustment weight coefficient in the Kth round of adjustment, The first word segmentation unit Probability of occurrence in the same syllable vocabulary; According to the adjustment weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment is calculated; which is expressed as follows: ， Where, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit Adjustment weight coefficient in the Kth round of adjustment; According to the word frequency weight coefficient of the first word segmentation unit in the Kth round of adjustment, the word frequency of the first word segmentation unit in the K+1th round of adjustment is calculated; which is expressed as follows: ， Where, The first participle unit The word frequency in the K+1 round of adjustment, The first word segmentation unit The word frequency weight coefficient in the Kth round of adjustment, The first word segmentation unit The word frequency of .

6. The input method word frequency adjustment device according to claim 4, characterized in that: The word segmentation model is a neural network algorithm model based on language grammar rules, which is used to identify the grammatical components of corpus data or the parts of speech of corpus data phrases, and segment the corpus data according to the grammatical components or the parts of speech of corpus data phrases; The phonetic recognition model is a neural network algorithm model based on language grammar rules, which is used to recognize syllables in corpus data and perform phonetic annotation.

Images